Hazardous waste supervision system based on Internet of Things

Abstract

The invention relates to the technical field of hazardous waste supervision, in particular to a hazardous waste supervision system based on the Internet of Things, which comprises a sensing layer, a network layer, a platform layer and an application layer, the sensing layer collects the whole process data of production, storage, transportation and disposal of hazardous wastes through a multi-dimensional sensing group, and the data accuracy is improved in combination with a dynamic threshold calibration algorithm; the network layer adopts hybrid networking of edge computing, 5G / NB-IoT and satellite communication, and realizes real-time transmission and non-tampering storage of data in cooperation with block chain nodes; the platform layer integrates a digital twin engine, an intelligent path planning module and a multi-department collaborative early warning system, and constructs a hazardous waste full life cycle virtual simulation and risk prediction model; the application level provides differentiated functions for supervision departments, waste production enterprises, transportation units and disposal institutions. According to the invention, real-time, precise, closed-loop and intelligent hazardous waste supervision is realized, the supervision efficiency is greatly improved, and the environmental risk is reduced.

Description

Technical Field

[0001] This invention relates to the field of hazardous waste management technology, specifically to an Internet of Things-based hazardous waste management system. Background Technology

[0002] Hazardous waste refers to solid waste listed in the National Hazardous Waste Inventory or identified as having hazardous characteristics according to national hazardous waste identification standards and methods. It possesses characteristics such as corrosivity, toxicity, flammability, reactivity, or infectivity. Improper supervision can cause serious harm to the ecological environment and human health. With increasingly frequent industrial production and social activities, the amount of hazardous waste generated continues to grow, highlighting the increasing complexity and urgency of its supervision.

[0003] The existing hazardous waste management model mainly relies on manual reporting, periodic inspections, and paper-based record management, which presents numerous unresolved problems. First, data collection and monitoring suffer from significant lag. Under the manual reporting system, information such as the generation, storage status, and transportation routes of hazardous waste must be proactively reported by enterprises. Regulatory authorities struggle to obtain accurate data in real time, leading to frequent instances of concealment, omissions, and misreporting. Some enterprises, in an effort to reduce costs, even illegally dump or transfer hazardous waste, posing a significant threat to environmental safety. Second, existing IoT monitoring systems often use single sensors to collect data, failing to consider the impact of factors such as temperature, humidity, electromagnetic interference, and chemical corrosion in the hazardous waste storage and transportation environment on the accuracy of the sensor data. This results in a high false alarm rate, making it difficult for regulatory authorities to make accurate judgments based on the data. Furthermore, data storage often relies on centralized servers, posing a risk of data tampering and loss. In the event of an environmental incident, it is difficult to trace responsibility and clearly identify the responsible parties at each stage, including waste-generating enterprises, transportation units, and disposal agencies.

[0004] In terms of transportation supervision, existing technologies mostly rely on GPS positioning to record transportation routes, but they fail to incorporate real-time road conditions, restricted areas for hazardous waste transportation (such as water source protection areas and residential areas), and road capacity for intelligent route planning. This results in low transportation efficiency and an inability to promptly detect abnormalities such as route deviations and illegal stops. Furthermore, the lack of effective real-time monitoring methods during hazardous waste transportation makes it impossible to provide timely warnings about the sealing status of transport vehicles and potential hazardous waste leaks. In the event of a leak, environmental pollution can easily spread.

[0005] In terms of full life-cycle supervision, existing systems mostly focus on the supervision of single links, lacking data integration across all stages of hazardous waste generation, storage, transportation, and disposal, forming "information silos." Regulatory authorities cannot fully grasp the flow of hazardous waste, making it difficult to achieve closed-loop supervision throughout the entire process. Furthermore, existing systems lack predictive supervision capabilities, only able to react passively after an accident occurs, unable to anticipate risks in advance through data analysis, such as exceeding the storage period for hazardous waste or insufficient disposal capacity, leading to a reactive approach to supervision.

[0006] In terms of emergency response, the existing regulatory system lacks a multi-departmental coordination mechanism. When emergencies such as hazardous waste leaks or fires occur, manual coordination among multiple departments, including environmental protection, emergency response, public security, and medical services, is required. This results in long response times, cumbersome procedures, and a risk of delaying optimal response times and expanding the scope of the accident's impact. Furthermore, emergency response plans are mostly generic templates, failing to develop personalized disposal recommendations based on the specific type of hazardous waste and the accident site environment. This leads to a lack of targeted measures and ineffective disposal.

[0007] In summary, existing hazardous waste monitoring technologies suffer from problems such as data lag, low accuracy, unreliable traceability, lack of a closed-loop system covering the entire life cycle, and inefficient emergency response. There is an urgent need for a technical solution that can achieve real-time, precise, closed-loop, and intelligent monitoring to improve the level of hazardous waste monitoring and protect the ecological environment and human health. Summary of the Invention

[0008] This invention aims to address the technical problems existing in the current hazardous waste supervision, such as data lag, low accuracy of sensor data, unreliable traceability system, fragmented supervision throughout the entire life cycle, insufficient emergency response coordination, and lack of predictive supervision capabilities. It provides an Internet of Things-based hazardous waste supervision system to achieve closed-loop supervision of hazardous waste from generation to disposal, thereby improving supervision efficiency and accuracy.

[0009] The technical solution adopted by this invention to solve its technical problem is: an Internet of Things-based hazardous waste monitoring system, comprising a sensing layer, a network layer, a platform layer, and an application layer; The sensing layer includes a multi-dimensional sensor group deployed at hazardous waste generation sites, storage warehouses, transport vehicles, and disposal terminals. The multi-dimensional sensor group includes a gas sensor, a temperature and humidity sensor, a liquid level / weight sensor, a GPS / BeiDou positioning module, a three-axis accelerometer, and a sealing status sensor. It is used to collect physicochemical parameters, location information, and status data of hazardous waste throughout its entire life cycle. The sensing layer also has a built-in dynamic threshold calibration unit that performs real-time calibration of the sensor data based on the LSTM algorithm. The network layer includes edge computing nodes, a hybrid communication module, and blockchain consensus nodes. The edge computing nodes preprocess the sensor data and perform initial anomaly detection. The hybrid communication module adopts a redundant design of 5G, NB-IoT, and satellite communication. The blockchain consensus nodes include regulatory department nodes, enterprise nodes, and third-party testing nodes to realize data on-chain evidence storage. The platform layer includes a digital twin engine, a blockchain traceability module, an intelligent route planning module, and a multi-department collaborative early warning system. The digital twin engine constructs a digital twin model of the entire life cycle of hazardous waste. The blockchain traceability module achieves tamper-proof data traceability based on a consortium blockchain architecture. The intelligent route planning module combines GIS maps and real-time traffic conditions to generate the optimal transportation route. The multi-department collaborative early warning system automatically matches the responsible parties and triggers tiered early warnings. The application layer includes a regulatory terminal, an enterprise terminal, a transportation terminal, and an emergency terminal. Each terminal interacts with data and calls functions through the platform layer, forming a closed-loop supervision process.

[0010] Specifically, the calibration process of the dynamic threshold calibration unit is as follows: collect sensor data and environmental interference parameters, establish a sensor data-interference parameter mapping model through LSTM algorithm, adjust the sensor threshold range in real time, eliminate abnormal interference data, and the data accuracy after calibration is not less than 98%.

[0011] Specifically, the blockchain consensus node adopts the PBFT consensus mechanism, and the block data includes hazardous waste classification information, sensor data, transportation trajectory data, disposal result data, and signature information of responsible entities at each stage. The block generation interval does not exceed 30 seconds.

[0012] Specifically, the digital twin engine acquires physical scene data through 3D laser scanning, constructs a static model by combining it with BIM technology, and updates the dynamic model in real time by accessing the perception layer data, thereby realizing the prediction of hazardous waste inventory, simulation of risk points, and optimization of disposal processes.

[0013] Specifically, the intelligent route planning module is based on the A* algorithm, incorporates data on hazardous waste transportation restricted areas, road carrying capacity, and real-time traffic congestion, and dynamically adjusts the transportation route, with a route deviation warning response time of no more than 10 seconds.

[0014] Specifically, the multi-department collaborative early warning system is equipped with a three-level early warning mechanism. The first-level early warning triggers a joint effort among environmental protection, emergency response, and public security departments. The second-level early warning is pushed to the regulatory authorities in the jurisdiction and the enterprise leaders. The third-level early warning only sends rectification reminders to the enterprises. The early warning information includes the location and type of risk and disposal suggestions.

[0015] Specifically, the multi-dimensional sensing group on the transport vehicle also includes a video monitoring module and an electronic seal sensor. The electronic seal sensor is linked to the door lock of the transport vehicle and automatically triggers positioning tracking and video recording when the seal is abnormal.

[0016] Specifically, the platform layer also includes a data encryption module, which uses the SM4 symmetric encryption algorithm to encrypt the transmitted data and the SM2 asymmetric encryption algorithm to authenticate the node identity.

[0017] Specifically, the enterprise side of the application layer supports online declaration of hazardous waste, automatic generation of ledgers, and compliance self-check functions, and automatically links to the updated classification standards of the national hazardous waste list.

[0018] Specifically, the sensors in the sensing layer are designed to be waterproof, corrosion-resistant, and explosion-proof, with a protection level of no less than IP67, making them suitable for the harsh environment of hazardous waste storage and transportation.

[0019] The beneficial effects of this invention are: Real-time monitoring: Through multi-dimensional sensor groups and hybrid communication networks, data on the entire life cycle of hazardous waste is collected in real time, allowing regulatory authorities to keep abreast of the status of hazardous waste, solving the problem of lagging data in traditional monitoring and improving the timeliness of monitoring.

[0020] Improve data accuracy: By adopting multi-sensor fusion and LSTM dynamic threshold calibration algorithm, abnormal data caused by environmental interference is eliminated, resulting in high data accuracy. This solves the problem of high false alarm rate of sensor data in existing systems and provides reliable data support for regulatory decision-making.

[0021] Ensuring traceability reliability: Based on a blockchain consortium blockchain architecture, the entire process of hazardous waste data is stored in an immutable and verifiable manner. Each batch of hazardous waste can be queried for information throughout the entire chain through a unique traceability code, clarifying the responsible parties at each stage and solving the problems of difficulty in traceability and unclear responsibilities.

[0022] Establish closed-loop supervision: break down data barriers in all aspects of hazardous waste generation, storage, transportation, and disposal to achieve closed-loop supervision throughout the entire life cycle, avoid information silos, and ensure that hazardous waste is controllable throughout the entire process.

[0023] Optimize emergency response efficiency: Establish a multi-departmental collaborative early warning mechanism and a personalized response plan recommendation function. The response time for a Level 1 early warning will not exceed 3 minutes, significantly shortening the emergency response cycle and reducing the scope of the accident's impact.

[0024] Possesses predictive regulatory capabilities: Through digital twin models, it can predict hazardous waste inventory and simulate risk points, anticipate regulatory risks in advance, and realize the transformation from passive to proactive regulation. Attached Figure Description

[0025] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] Figure 1 This is a diagram illustrating the architecture of the IoT-based hazardous waste monitoring system of this invention. Figure 2 This is a schematic diagram of the deployment of the multi-dimensional sensing group in the perception layer of the present invention; Figure 3 This is a flowchart illustrating the modeling process of the digital twin engine of this invention. Detailed Implementation

[0027] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0028] like Figures 1-3 As shown, the present invention provides an Internet of Things-based hazardous waste monitoring system, comprising a sensing layer, a network layer, a platform layer, and an application layer, with each layer working collaboratively. The perception layer is the core of data acquisition, deployed at hazardous waste generation points, storage warehouses, transport vehicles, and disposal terminals. It employs a multi-dimensional sensor array to achieve full-parameter coverage. This array includes gas sensors (detecting the concentration of toxic and harmful gases such as VOCs and heavy metal vapors), temperature and humidity sensors (monitoring the temperature and humidity of the storage and transport environment), liquid level / weight sensors (recording the amount of hazardous waste stored and transported), a GPS / BeiDou dual-mode positioning module (accuracy ≤5 meters), a three-axis accelerometer (monitoring for illegal dumping or severe collisions in transport vehicles), a sealing status sensor (detecting the sealing integrity of transport compartments and storage containers), and a video monitoring module (capturing real-time footage). The perception layer incorporates a dynamic threshold calibration unit, using an LSTM (Long Short-Term Memory) algorithm for real-time calibration of sensor data. The specific process is as follows: first, raw sensor data and environmental interference parameters (such as temperature, humidity, and electromagnetic intensity) are collected; then, an LSTM algorithm is used to train and establish a sensor data-interference parameter mapping model; finally, the sensor threshold range is adjusted in real-time based on the environmental interference parameters to eliminate abnormal data caused by environmental interference, ensuring that the accuracy of the sensor data is no less than 98%. The sensor is designed to be waterproof, corrosion-resistant, and explosion-proof, with a protection level of no less than IP67, making it suitable for the harsh environments of hazardous waste storage and transportation.

[0029] The network layer is responsible for data transmission and secure storage, employing an architecture of "edge computing + hybrid communication + blockchain". Edge computing nodes are deployed at waste-generating enterprises, transport vehicles, and disposal facilities, using NVIDIA Jetson Xavier NX edge computing modules to preprocess the raw data collected by the perception layer (such as data filtering, format conversion, and initial anomaly detection), reducing the amount of data transmitted and lowering the computational pressure on the platform layer. The preprocessing latency is no more than 50 milliseconds. The hybrid communication module adopts a redundant design of 5G, NB-IoT, and satellite communication: in areas with good network coverage, such as cities and industrial parks, 5G communication is used to achieve high-speed, low-latency data transmission; in remote areas or underground warehouses with weak network signals, NB-IoT is used to achieve low-power data transmission; in scenarios without terrestrial network coverage, such as cross-provincial transportation and maritime transportation, satellite communication is used to ensure continuous data transmission and ensure 100% data transmission coverage. The blockchain consensus nodes adopt a consortium blockchain architecture, including regulatory department nodes (environmental protection, emergency response, public security), enterprise nodes (waste-generating enterprises, transport units, disposal facilities), and third-party testing nodes, using the PBFT consensus mechanism, with a block generation interval of no more than 30 seconds. The block data includes hazardous waste classification information, sensor-collected data, transportation trajectory data, disposal result data, and signature information of responsible parties at each stage, ensuring that the data is tamper-proof and traceable. The network layer also includes a data encryption module, using the SM4 symmetric encryption algorithm to encrypt transmitted data and the SM2 asymmetric encryption algorithm to authenticate node identities, ensuring the security of data transmission and storage.

[0030] The platform layer is the core processing unit of the system, integrating a digital twin engine, a blockchain traceability module, an intelligent path planning module, and a multi-department collaborative early warning system to achieve functions such as data processing, modeling and analysis, and early warning response. The digital twin engine acquires 3D data of physical scenes such as hazardous waste generation points, storage warehouses, and disposal facilities through 3D laser scanning, and constructs a static model using BIM technology. It also receives dynamic data collected in real time from the perception layer (such as hazardous waste inventory, environmental parameters, and transportation trajectories) to update the dynamic model, forming a digital twin model of the entire lifecycle of hazardous waste. This model can predict hazardous waste inventory (based on historical generation and current storage levels, predicting the storage status for the next 7 days), simulate risk points (such as simulating the diffusion path after a leak), and optimize the disposal process (optimizing disposal process parameters through virtual simulation). The blockchain traceability module, based on a consortium blockchain architecture, assigns a unique traceability code to each batch of hazardous waste (using an 18-digit coding rule of "enterprise code + generation date + classification code + batch number"). The traceability code is bound to sensors, transport vehicles, and disposal equipment, enabling the entire process of hazardous waste from generation to disposal to be stored and verified on the blockchain. Regulatory authorities and enterprises can use traceability codes to query data across the entire chain of hazardous waste, including its generation, storage status, transportation trajectory, and disposal results, ensuring that responsibility is traceable.

[0031] The intelligent path planning module is based on the A* algorithm and incorporates restricted areas for hazardous waste transportation (such as water source protection areas, residential areas, and around schools), road carrying capacity (for heavy metal and corrosive hazardous waste), and real-time traffic congestion data to generate the optimal transportation path for the transport vehicle. When there are changes in road conditions, road closures, etc. during the transportation process, the module adjusts the path in real time and pushes it to the transport terminal, and the path deviation warning response time does not exceed 10 seconds. At the same time, the module can also plan the path passing through the emergency rescue points according to the characteristics of the hazardous waste (such as flammable, explosive, toxic) to ensure that emergencies can be handled in a timely manner. The multi-department collaborative warning system sets up a three-level warning mechanism: the first-level warning (such as a large leakage of hazardous waste, fire and explosion) triggers the joint action of multiple departments including environmental protection, emergency, public security, and medical, and automatically pushes the accident location, type of hazardous waste, and disposal suggestions to the terminals of each department; the second-level warning (such as storage expiration, slight path deviation) is pushed to the regulatory department of the jurisdiction and the person in charge of the enterprise, and requires a feedback on the processing results within 2 hours; the third-level warning (such as slight over-standard of temperature and humidity) only sends a rectification reminder to the enterprise and requires the rectification to be completed within 24 hours. The warning system can also automatically match the responsible entity according to the type and risk level of the hazardous waste to avoid shirking of responsibility.

[0032] At the application level, differentiated functions are provided to different user groups, including regulatory, enterprise, transportation, and emergency response terminals. Each terminal interacts and calls functions via web and mobile apps. The regulatory terminal, for environmental protection and emergency response regulatory departments, provides functions such as visualization of hazardous waste's entire lifecycle data, real-time monitoring, statistical analysis, and investigation of violations. Regulatory personnel can view data on hazardous waste generation, storage, transportation routes, and disposal status within their jurisdiction through a visual interface, generating monthly / quarterly regulatory reports; they can view the real-time status of key enterprises and key transportation routes through the real-time monitoring function, and directly send rectification notices to enterprises upon detecting anomalies; and they can understand the trends in hazardous waste generation and regional distribution characteristics through statistical analysis functions, providing data support for regulatory decisions. The enterprise terminal, for waste-generating enterprises and disposal institutions, provides functions such as online hazardous waste declaration, automatic ledger generation, compliance self-inspection, and equipment management. Enterprises can submit hazardous waste generation plans and transfer applications online. The system automatically generates electronic ledgers based on data from the sensing layer, eliminating the need for manual data entry and reducing workload. The compliance self-check function automatically identifies compliance risks in storage and disposal processes by comparing data against the national hazardous waste list and environmental regulations. The equipment management function monitors the operating status of sensors, storage devices, and disposal equipment in real time, promptly identifying and reporting equipment malfunctions. The transportation end provides intelligent route navigation, real-time monitoring, anomaly reporting, and transportation certificate generation for transportation units. Drivers can obtain the optimal transportation route through the intelligent route navigation function and receive real-time route adjustment notifications. The real-time monitoring function allows viewing data such as the location, speed, and sealing status of the transport vehicle; anomalies (such as seal damage or gas leakage) can be reported with one click. After transportation is completed, the system automatically generates a transportation certificate containing information such as the transportation trajectory, time, and responsible person, serving as the basis for handover. The emergency response end provides accident location, risk assessment, disposal plan recommendations, and resource scheduling for emergency rescue teams. After an accident occurs, the system automatically locates the accident site, assesses the scope of the accident's impact and risk level based on a digital twin model, recommends personalized disposal solutions (such as leakage disposal methods and protective measures) according to the type of hazardous waste and the accident scenario, and views the distribution of surrounding emergency rescue teams, equipment, and materials through the resource scheduling function to achieve rapid dispatch.

[0033] Example 1: Hazardous Waste Monitoring System in Industrial Parks This embodiment is applied to a chemical industrial park with 32 waste-generating enterprises, mainly producing hazardous waste such as waste acid, waste alkali, and waste containing heavy metals, with an annual production of approximately 5,000 tons. The park involves 12 storage warehouses, 28 transport vehicles, and 2 disposal facilities. The specific implementation of the IoT-based hazardous waste monitoring system in this embodiment is as follows: (a) Deployment of the perception layer Multi-dimensional sensor groups were deployed in the waste-generating workshops and storage warehouses of 32 waste-generating enterprises: In each production line of the waste-generating workshop, one gas sensor (model: MQ-138, detecting VOCs, hydrogen sulfide, etc., measurement range 0-1000ppm, accuracy ±5%FS) and one temperature and humidity sensor (model: SHT30, measurement range -40~125℃, 0~100%RH, accuracy ±0.3℃, ±2%RH) were deployed; in the storage warehouses, sensors were deployed at intervals of 50m². 2 Deploy one gas sensor and one temperature and humidity sensor. Install one liquid level / weight sensor on each storage tank / container (liquid level sensor model: YL-70, measuring range 0-5m, accuracy ±1mm; weight sensor model: YZC-320, measuring range 0-500kg, accuracy ±0.1%FS). Install one video monitoring module (model: Hikvision DS-2CD3T46WD-I3, resolution 1080P, supports night vision) at the warehouse entrance and exit.

[0034] One gas sensor, one temperature and humidity sensor, one sealing status sensor (model: FS-01, detecting sealing pressure, measurement range 0-1MPa, accuracy ±0.01MPa), one GPS / BeiDou dual-mode positioning module (model: UBLOXNEO-M8N, positioning accuracy ≤3 meters, update frequency 1Hz), one triaxial accelerometer (model: ADXL345, measurement range ±16g, accuracy ±0.01g), and one video monitoring module are installed in the compartments of 28 transport vehicles; one edge computing node (NVIDIA Jetson Xavier NX) is installed in the driver's cab for data preprocessing; and an electronic seal sensor (model: FQ-200, supporting NFC unlocking, with unlocking records automatically uploaded) is installed at the compartment door locks.

[0035] One gas sensor, one temperature and humidity sensor, one liquid level sensor, and one video monitoring module were deployed in the treatment workshop and wastewater treatment station of the two treatment facilities, respectively; one equipment operation status sensor (model: PT100, measuring equipment temperature, range 0-500℃, accuracy ±0.5℃) was installed on the treatment equipment.

[0036] All sensors are designed with IP68 protection rating and are coated with anti-corrosion coating to adapt to the corrosive environment of chemical industrial parks. The dynamic threshold calibration unit is integrated into the sensor control module and calibrates the data in real time through the LSTM algorithm, with a calibration cycle of 10 seconds / cycle.

[0037] (ii) Network layer deployment One blockchain consensus node server (configuration: Intel Xeon Gold 6248, 32GB RAM, 2TB SSD) is deployed at the park management center as a node for regulatory authorities. One edge computing node / blockchain node is deployed each for 32 waste-generating companies, 28 transport vehicles, and 2 disposal facilities (transport vehicles use vehicle-mounted edge computing nodes, others use fixed nodes), forming a consortium blockchain network. The blockchain uses the PBFT consensus mechanism, with a block generation interval of 20 seconds. Block data includes company codes, hazardous waste classification codes, sensor data, timestamps, and responsible person signatures.

[0038] The communication network adopts a hybrid architecture of "5G+NB-IoT": waste-generating enterprises, storage warehouses, and disposal institutions within the park achieve data transmission through the park's 5G base stations, with a transmission rate of ≥100Mbps and a latency of ≤20ms; transport vehicles use 5G communication when traveling within the park and on urban roads, and NB-IoT communication in remote areas to ensure continuous data transmission. Data encryption uses the SM4 symmetric encryption algorithm with a key length of 128 bits; node authentication uses the SM2 asymmetric encryption algorithm, and each node is assigned a unique digital certificate.

[0039] (III) Platform Layer Deployment The platform is deployed on a cloud server cluster in the park (configuration: 8 Intel Xeon Platinum 8275CL servers, each with 64GB of memory and 4TB SSD, managed by Kubernetes cluster), integrating a digital twin engine, a blockchain traceability module, an intelligent path planning module, and a multi-department collaborative early warning system.

[0040] Digital Twin Engine Modeling Process: First, a 3D laser scanner (model: Faro Focus S70) is used to perform 3D scanning of the waste-generating workshops of 32 enterprises, 12 storage warehouses, and the disposal workshops of 2 disposal facilities within the park, acquiring point cloud data. MeshLab software is used for point cloud preprocessing (denoising, registration, and simplification), and the data is imported into Revit software combined with BIM technology to construct a static 3D model. Dynamic data from the perception layer (sensor data, positioning data, and equipment operation data) is accessed in real-time via API interfaces, and the Unity3D engine is used to dynamically update the model, generating a digital twin model of the entire lifecycle of hazardous waste in the park. This model can display the amount of hazardous waste generated, stored, transported, and disposed of by each enterprise in real time, supporting zooming, panning, and other operations. A predictive model trained based on historical data can predict the hazardous waste storage status of each warehouse for the next 7 days, automatically triggering an alert when the predicted storage volume exceeds 80% of the warehouse capacity.

[0041] The blockchain traceability module assigns a unique 18-digit traceability code to each batch of hazardous waste (format: 4-digit company code + 8-digit production date + 3-digit classification code + 3-digit batch number). For example, "000120240520003001" represents the 001st batch of Class 003 hazardous waste produced by company number 0001 on May 20, 2024. The traceability code is bound to the hazardous waste container via RFID tags. Sensors, transport vehicles, and disposal equipment are all associated with the traceability code, enabling real-time data uploading to the blockchain. Regulatory personnel can enter the traceability code through the regulatory terminal to query the entire chain of data for that batch of hazardous waste, including its production time, quantity, composition test report, storage location, transport vehicle information, transport trajectory, disposal time, and disposal results.

[0042] The intelligent route planning module, based on the A* algorithm, imports a GIS map of the park and surrounding areas (including road information, hazardous waste restricted areas, and emergency rescue point distribution), and combines it with real-time traffic data provided by the transportation department to generate the optimal route for transport vehicles. For example, when transporting waste acid from company number 0001 to disposal facility number 001, the system automatically avoids water source protection areas and residential areas, selects routes with a road carrying capacity of ≥20 tons, and passes through 2 emergency rescue points. When congestion occurs on a certain section of the road during transportation, the system adjusts the route in real time and pushes the new route through the transportation terminal APP, with a route adjustment response time of ≤8 seconds.

[0043] The multi-department collaborative early warning system is equipped with three levels of warning thresholds: Level 1 warning (gas concentration exceeds the national limit by 1.2 times, pressure detected by the sealing status sensor is below 0.05MPa, impact acceleration detected by the triaxial accelerometer is ≥8g), Level 2 warning (temperature and humidity exceed the set range by ±5%, storage time exceeds 80% of the specified period, path deviation is ≥500 meters), and Level 3 warning (gas concentration exceeds the national limit by 1.0-1.2 times, temperature and humidity exceed the set range by ±3%-±5%). When the gas sensor of the transport vehicle detects a waste acid leak and the concentration reaches 1.5 times the national limit, a Level 1 warning is triggered. The system automatically pushes the accident location (latitude and longitude: 30°15′23″N, 120°08′45″E), hazardous waste type (waste acid, pH≤2), estimated leakage amount (based on liquid level sensor data, approximately 50kg), and disposal recommendations (neutralization with lime, evacuation radius 50 meters) to the terminals of environmental protection, emergency response, public security, and medical departments, and simultaneously pushes the information to the transportation and disposal ends. The emergency response time is 2 minutes and 30 seconds.

[0044] (iv) Application layer deployment The monitoring platform utilizes a combination of web and mobile app. The web interface is deployed on the office computers of the park's environmental supervision department, while the mobile app is installed on the mobile phones of supervisors (supporting Android and iOS systems). The monitoring platform displays real-time statistics on the generation, storage, transportation, and disposal of hazardous waste in the park, showing monthly trends in chart format. The real-time monitoring page allows users to view the real-time status of each company's storage warehouses and transport vehicles. Clicking on a specific transport vehicle displays its location, speed, gas concentration inside the vehicle, and sealing status. The statistical analysis function generates park hazardous waste classification statistical reports and company compliance assessment reports, providing support for regulatory decision-making.

[0045] The enterprise-side interface is installed on the office computers and mobile phones of 32 waste-generating enterprises and 2 disposal institutions. Enterprises can submit hazardous waste generation plans online and upload component testing reports through the enterprise-side interface. The system automatically generates electronic ledgers based on the data from the perception layer, including information such as generation date, generation quantity, storage location, transfer time, and receiving unit, and can be exported in PDF format for filing. The compliance self-check function compares the hazardous waste against the "National Hazardous Waste List (2021 Edition)" to automatically check whether the hazardous waste classification is accurate, whether the storage time has exceeded the expiration date, and whether the disposal unit has the qualifications. If violations are found, the system will automatically remind enterprises to rectify them.

[0046] The transportation app is installed on the drivers' mobile phones and in-vehicle terminals of 28 transport vehicles. Drivers receive transportation tasks (including hazardous waste batches, origin, destination, and optimal route) through the app. During transportation, the app displays the vehicle's location, speed, and environmental parameters inside the vehicle in real time. When an abnormality occurs (such as a broken seal), the app issues an audible and visual alarm, and the driver can report it to the platform with one click. After transportation is completed, the driver and the disposal agency staff confirm the handover through the app, and the system automatically generates a transportation voucher, which includes the transportation trajectory, time, and the signature of the responsible person, serving as the basis for settlement.

[0047] The emergency terminal is installed on the command vehicle of the emergency rescue team in the park and on the mobile phones of the rescue personnel. When a Level 1 warning is issued, the emergency terminal will automatically display the accident location, type of hazardous waste, risk level, disposal suggestions, and distribution of surrounding emergency resources (such as the nearest fire hydrant and emergency material warehouse). Commanders can use the emergency terminal to dispatch rescue teams and view the rescue progress in real time.

[0048] (V) Operational Results After six months of operation, the system in this embodiment achieved the following results: the accuracy rate of hazardous waste declaration increased from 85% to 99.5%, data transmission latency was ≤30ms, and the accuracy rate of abnormal warning was 98.8%; three hazardous waste leakage accidents were successfully warned and handled, and the emergency response time was shortened from 30 minutes to less than 3 minutes; there were zero cases of illegal transfer of hazardous waste, and the response time for traceability queries was ≤2 seconds; the workload of manual verification by regulatory authorities was reduced by 70%, and regulatory efficiency was greatly improved.

[0049] Example 2: Medical Hazardous Waste Monitoring System This embodiment is applied to the medical system of a certain province, covering 10 tertiary hospitals, 20 secondary hospitals, 50 community hospitals, and 3 medical hazardous waste disposal facilities. The system mainly generates infectious waste, sharps waste, pathological waste, pharmaceutical waste, and chemical waste, with an annual production of approximately 3,000 tons. The specific implementation of the IoT-based hazardous waste monitoring system in this embodiment is as follows: (a) Deployment of the perception layer Multi-dimensional sensor arrays are deployed in various hospital departments (such as operating rooms, laboratories, and infectious disease departments) and medical waste storage areas: Each department's medical waste collection point is equipped with one sealing status sensor (model: FS-02, adapted to medical waste transfer boxes, detecting whether the seal is intact), one weight sensor (model: YZC-310, measurement range 0-100kg, accuracy ±0.1%FS), and one infectious risk sensor (model: GR-01, detecting bacterial concentration, measurement range 0-1000 CFU / m³). 3 Accuracy ±10 CFU / m 3 The infectious disease department also installed an additional gas sensor (model: MQ-137, to detect gases such as ammonia and formaldehyde).

[0050] Medical waste temporary storage rooms are set at 30m² 2 Deploy one temperature and humidity sensor (model: SHT35, measurement range -20~85℃, 0~100%RH, accuracy ±0.2℃, ±1.8%RH), one gas sensor, and one video monitoring module (model: Dahua DH-IPC-HFW3249M-I1, supports motion detection). Each temporary storage cabinet is equipped with one liquid level sensor (for liquid medical waste) and one electronic seal sensor (model: FQ-201, supports password unlocking and uploads unlocking records).

[0051] In the high-temperature sterilization workshop and incineration workshop of the three disposal facilities, one temperature and humidity sensor, one gas sensor, and one equipment operation status sensor (model: DS18B20, measuring equipment temperature, range -55~125℃, accuracy ±0.5℃) were deployed. In the wastewater treatment station, one pH sensor (model: PH-200, measuring range 0-14pH, accuracy ±0.01pH) was deployed.

[0052] Each of the 40 transport vehicles is equipped with one gas sensor, one temperature and humidity sensor, one sealing status sensor, one GPS / BeiDou dual-mode positioning module (model: UBLOX NEO-7M, positioning accuracy ≤5 meters), one triaxial accelerometer (model: MPU6050, measurement range ±8g, accuracy ±0.02g), and one video monitoring module. An edge computing node (NVIDIA Jetson Nano) is installed in the driver's cab, and an electronic seal sensor is installed at the door lock of the vehicle compartment.

[0053] For sharp waste (such as needles and blades), a sharps counting sensor (model: JS-01, which counts via infrared sensing with an accuracy of ±1) is installed on the collection container; all sensors have a protection rating of no less than IP67, making them suitable for hospital cleaning and disinfection environments; the dynamic threshold calibration unit is based on the LSTM algorithm and optimizes the calibration model to address interference factors such as disinfectant evaporation in the hospital environment, with a calibration cycle of 5 seconds per calibration.

[0054] (ii) Network layer deployment One main blockchain node server (configuration: Intel Xeon Gold 6348, 64GB RAM, 4TB SSD) was deployed at the Provincial Health Commission as a regulatory node. One blockchain node was also deployed at each of the following locations: 10 tertiary hospitals, 20 secondary hospitals, 50 community hospitals, 40 transport vehicles, and 3 disposal facilities. Hospitals and disposal facilities used fixed nodes, while transport vehicles used vehicle-mounted nodes, forming a consortium blockchain network. The blockchain uses the PBFT consensus mechanism, with a block generation interval of 15 seconds. Block data includes hospital codes, medical waste classification codes, sensor data, disinfection records, transportation information, and disposal results.

[0055] The communication network adopts a hybrid architecture of "5G + LAN": sensors inside the hospital connect to the hospital's blockchain node via the LAN, and then upload data to the platform layer via the 5G network; transport vehicles use 5G communication, automatically switching to NB-IoT communication in remote areas. Data encryption uses the SM4 symmetric encryption algorithm, and node identity authentication uses the SM2 asymmetric encryption algorithm to ensure the privacy and security of medical waste data (including indirect patient information).

[0056] (III) Platform Layer Deployment The platform is deployed on a provincial government cloud server cluster (configuration: 10 Intel Xeon Platinum 8375C servers, each with 128GB of memory and 8TB of SSD, managed using Kubernetes cluster), integrating a digital twin engine, a blockchain traceability module, an intelligent path planning module, and a multi-department collaborative early warning system.

[0057] Digital Twin Engine Modeling Process: A 3D laser scanner (Leica BLK360) is used to perform 3D scanning of the medical waste storage rooms and disposal workshops of various hospitals to obtain point cloud data. AutoCAD software is used for point cloud processing, and the data is imported into 3ds Max to construct a static 3D model. Dynamic data from the perception layer is accessed in real-time via API interfaces, and Unreal Engine 4 is used to dynamically update the model, generating a digital twin model of the entire lifecycle of hazardous medical waste. This model can display the amount of medical waste generated, its storage status, transportation trajectory, and disposal progress in real time for each hospital. For infectious waste, its transmission path can be simulated to assess pollution risks. Based on historical data, the model predicts the trend of medical waste generation in each hospital. When the predicted storage volume in the temporary storage room exceeds 70% of its capacity, the model automatically alerts the transportation unit to arrange for transfer.

[0058] The blockchain traceability module assigns a unique 18-digit traceability code to each batch of medical waste (format: 4-digit hospital code + 2-digit department code + 6-digit production date + 3-digit classification code + 3-digit batch number). For example, "00020120240610001002" represents batch 002 of category 001 (infectious waste) generated on June 10, 2024, by department 01 of hospital number 0002. The traceability code is bound to the medical waste transfer container via RFID tags. Sensors, transport vehicles, and disposal equipment are linked to the traceability code, and the data is uploaded to the blockchain in real time. Health commissions and environmental protection departments can use the traceability code to query the entire chain of data for that batch of medical waste, including the generating department, production time, weight, sealing status, transport vehicle information, disinfection records, and disposal method.

[0059] The intelligent route planning module, based on the A* algorithm, imports a province-wide GIS map, marking restricted areas for medical waste transportation (such as areas around schools, kindergartens, and food processing plants), the locations of medical waste disposal facilities, and the distribution of emergency rescue points. Combined with real-time traffic data, it generates the optimal route for transport vehicles. For example, when transporting infectious waste from a tertiary hospital (number 0005) to a disposal facility (number 002), the system automatically avoids densely populated areas and selects the route with the shortest travel time that passes through emergency rescue points. Transport vehicles must complete the transportation within a specified time window (e.g., 2-6 AM), and the system monitors the transportation time in real time, automatically issuing warnings if the time limit is exceeded.

[0060] The multi-department collaborative early warning system sets three levels of early warning thresholds: Level 1 warning (bacterial concentration detected by infectious risk sensor ≥500 CFU / m³) 3 Level 1 warning (damaged seal, impact acceleration of transport vehicle ≥6g), Level 2 warning (storage time exceeds 48 hours, temperature and humidity exceed standard ±5%, route deviation ≥300 meters), Level 3 warning (bacterial concentration 300-500 CFU / m³). 3 Temperature and humidity exceeding the standard (±3%-±5%). When the infectious risk sensor in a community hospital's temporary storage room detected a bacterial concentration of 600 CFU / m³... 3 When a Level 1 warning is triggered, the system automatically pushes the accident location, medical waste type, risk level, and disposal suggestions (such as immediate disinfection and isolation storage) to the Health Commission, environmental protection department, emergency department, and disposal agency. At the same time, it reminds the hospital to take prevention and control measures. The emergency response time is ≤2 minutes.

[0061] (iv) Application layer deployment The regulatory platform includes monitoring terminals from the Health Commission and the Environmental Protection Department, utilizing both web and mobile apps. The platform displays real-time data on the province's medical waste generation, classification statistics, transportation volume, and disposal volume, generating monthly regulatory reports. It also monitors the status of temporary storage rooms and transport vehicles in hospitals in real time, allowing users to view video surveillance footage. For infectious waste, a dedicated monitoring module tracks its entire process, ensuring that disinfection and disposal comply with regulations.

[0062] The enterprise-side (hospital-side and disposal facility-side) is installed in the logistics management departments of each hospital and the operation departments of disposal facilities. The hospital-side supports functions such as online medical waste declaration, departmental ledger generation, and temporary storage room management. The system automatically generates electronic ledgers based on data from the perception layer, including information such as department, amount generated, transfer time, and receiving unit, and can be connected to the hospital's HIS system. The disposal facility-side supports functions such as receiving confirmation, disposal process recording, and uploading disposal results. After disposal is completed, a disposal report is automatically generated and uploaded to the platform layer.

[0063] The transportation app is installed on the drivers' mobile phones and vehicle terminals of 40 transport vehicles. Drivers receive transfer tasks through the app and view the hospital location, temporary storage room information, and optimal route. During transportation, the app displays the vehicle location, bacterial concentration in the compartment, and sealing status in real time, and issues an alarm when there are abnormalities. After arriving at the treatment facility, the driver and staff confirm the handover through the app, and the system automatically generates a transportation voucher.

[0064] The emergency terminal is installed on the terminal equipment of the provincial emergency rescue team and emergency rescue teams in various cities and counties. After the first-level warning is triggered, the emergency terminal will automatically display the accident location, the type of medical waste, disposal suggestions and surrounding emergency resources (such as disinfection material warehouses and medical rescue teams). Commanders can use the emergency terminal to dispatch resources and monitor the disposal progress in real time.

[0065] (V) Operational Results After eight months of operation, the accuracy rate of medical waste declaration in this embodiment increased from 88% to 99.8%, the compliance rate of infectious waste disposal increased from 90% to 99.5%, and the accuracy rate of abnormal warning reached 99.2%. Two incidents of medical waste sealing damage were successfully handled without any spread of infection. The average temporary storage time of medical waste was shortened to less than 24 hours, and transportation efficiency was improved by 30%. The workload of manual verification by regulatory authorities was reduced by 80%, achieving precise and intelligent supervision of medical hazardous waste.

[0066] Example 3: Inter-provincial Hazardous Waste Transportation Supervision System This embodiment applies to a scenario involving the inter-provincial transportation of hazardous waste, involving 5 waste-generating enterprises in Province A (mainly producing waste catalysts, waste mineral oil, and chromium-containing waste) and 3 disposal facilities in Province B, with an annual inter-provincial transportation volume of approximately 2,000 tons. The transportation routes cover highways, national roads, provincial roads, and some remote sections. The specific implementation of the IoT-based hazardous waste monitoring system in this embodiment is as follows: (a) Deployment of the perception layer Multi-dimensional sensor arrays were deployed in the storage warehouses of five waste-generating enterprises in Province A: each warehouse was equipped with sensors for every 100m² of storage space. 2 One gas sensor (model: MQ-139, for detecting volatile gases from waste mineral oil, measurement range 0-500ppm, accuracy ±3%FS) and one temperature and humidity sensor (model: SHT40, measurement range -40~125℃, 0~100%RH, accuracy ±0.1℃, ±1.5%RH) are deployed. Each storage tank is equipped with one liquid level sensor (model: JL-80, measurement range 0-10m, accuracy ±2mm) and one weight sensor (model: YZC-520, measurement range 0-1000kg, accuracy ±0.05%FS). Video monitoring modules (model: Ezviz C6CN, supporting 360° rotation) and electronic seal sensors (model: FQ-300, supporting remote unlocking with hierarchical unlocking permission management) are installed at the warehouse entrance and exit.

[0067] In the cargo compartments of 15 inter-provincial transport vehicles (all dedicated to transporting hazardous materials), one gas sensor, one temperature and humidity sensor, one sealing status sensor (model: FS-03, adapted to the bumpy environment of long-distance transportation, detecting sealing pressure, range 0-2MPa, accuracy ±0.02MPa), one GPS / BeiDou dual-mode positioning module (model: UBLOX NEO-M9N, positioning accuracy ≤2 meters, update frequency 2Hz), one triaxial accelerometer (model: ADXL355, measurement range ±2g / ±4g / ±8g selectable, accuracy ±0.001g), one video monitoring module (supporting driving recorder function), and one satellite communication module (model: Iridium 9602, ensuring communication in areas without terrestrial network); an edge computing node (NVIDIA Jetson AGX Xavier) is installed in the vehicle cab for data preprocessing and local storage (1TB storage capacity, capable of caching 7 days of data); and one leak detection sensor (model: XL-01, detecting liquid leaks, response time ≤1 second) is installed at the bottom of the cargo compartment.

[0068] In the receiving area, storage warehouse, and disposal workshop of the three disposal facilities in Province B, one gas sensor, one temperature and humidity sensor, one weight sensor, and one equipment operation status sensor (model: PT1000, measuring equipment temperature, range 0-800℃, accuracy ±0.1℃) are deployed. Video monitoring modules and electronic seal sensors are installed in the receiving area.

[0069] All sensors are designed to be explosion-proof, making them suitable for long-distance transportation and complex road conditions. The dynamic threshold calibration unit optimizes the LSTM calibration model and adds air pressure parameter input to address the large temperature and air pressure changes during inter-provincial transportation. The calibration cycle is 5 seconds per cycle.

[0070] (ii) Network layer deployment One main blockchain node server (configuration: Intel Xeon Gold 6448Y, 48GB RAM, 3TB SSD) is deployed in each of the environmental protection departments of Province A and Province B as provincial-level regulatory nodes. One blockchain node is deployed each in each of the five waste-generating companies, 15 transport vehicles, and three disposal facilities. One backup blockchain node is deployed in each of the environmental protection departments of the provinces the waste is transported through (e.g., Province C and Province D), forming a cross-regional consortium blockchain network. The blockchain uses the PBFT consensus mechanism, with a block generation interval of 25 seconds. Block data includes information such as the waste-generating province code, company code, hazardous waste classification code, sensor data, transport trajectory, areas traversed, and disposal results.

[0071] The communication network adopts a hybrid architecture of "5G + NB-IoT + satellite communication": 5G communication is used when transport vehicles are traveling in cities and on highways, with a transmission rate of ≥200Mbps and a latency of ≤15ms; NB-IoT communication is used in areas with weak network signals, such as national and provincial highways; and satellite communication is used in remote sections, mountainous areas, and other areas without terrestrial network coverage to ensure continuous data transmission. Data encryption uses the SM4 symmetric encryption algorithm, cross-provincial data transmission uses VPN tunnel encryption, node authentication uses the SM2 asymmetric encryption algorithm, and each node is assigned a unique digital certificate jointly issued by the environmental protection departments of provinces A and B.

[0072] (III) Platform Layer Deployment The platform layer is deployed on the national-level hazardous waste supervision cloud platform (configuration: 15 Intel Xeon Platinum8470C servers, each with 256GB of memory and 16TB of SSD, using a distributed architecture), integrating a digital twin engine, a blockchain traceability module, an intelligent path planning module, and a multi-department collaborative early warning system.

[0073] Digital Twin Engine Modeling Process: A 3D laser scanner (Trimble X7) is used to perform 3D scanning of the storage warehouses of 5 waste-generating enterprises in Province A, the disposal workshops of 3 disposal institutions in Province B, and the environment along the main transportation routes to obtain point cloud data. CloudCompare software is used for point cloud processing, and the data is imported into Bentley MicroStation software combined with GIS technology to construct a static 3D model. Dynamic data (sensor data, positioning data, and road condition data) from the perception layer is accessed in real time via API interfaces. The ContextCapture engine is used to dynamically update the model, generating a digital twin model of the entire lifecycle of inter-provincial hazardous waste transportation. This model can display the real-time location and speed of transport vehicles, environmental parameters inside the vehicle compartment, and the status of hazardous waste. It simulates potential risks encountered during transportation (such as insufficient bridge load-bearing capacity, tunnel height restrictions, and the impact of severe weather) and proactively mitigates these risks. Based on historical transportation data and real-time road conditions, it predicts the arrival time of transportation with an error of no more than 30 minutes.

[0074] The blockchain traceability module assigns a unique 18-digit traceability code to each batch of hazardous waste transported across provinces (format: 2-digit province of origin code + 4-digit company code + 6-digit production date + 3-digit classification code + 3-digit batch number). For example, "10000820240705005003" represents batch 003 of category 005 (waste mineral oil) generated on July 5, 2024, by company number 0008 in province 10 (province A). The traceability code is bound to the hazardous waste container via RFID tags. Sensors, transport vehicles, and disposal equipment are linked to the traceability code, and data is uploaded to the blockchain in real time. Environmental protection departments in provinces A, B, and transit provinces can query the entire chain data of this batch of hazardous waste through the traceability code, achieving cross-regional collaborative supervision.

[0075] The intelligent route planning module, based on an improved A* algorithm, imports a national GIS map, marking restricted areas for hazardous waste transportation (such as water source protection areas, nature reserves, and areas surrounding key cultural relics protection units), road carrying capacity, tunnel height and width restrictions, and the distribution of emergency rescue points. Combined with real-time traffic data from transportation departments and weather data from meteorological departments (such as rainstorm and snowstorm warnings), it generates the optimal route for transport vehicles. For example, transporting waste mineral oil from company No. 0008 in province A to disposal facility No. 003 in province B, the system automatically avoids dangerous mountain roads and areas under rainstorm warnings, selecting a route primarily using highways and secondarily using national roads, passing through 3 emergency rescue points and 2 provincial checkpoints. During transportation, the system monitors road conditions in real time. If a section of road is closed due to construction, the system immediately adjusts the route and pushes the adjustment to the transportation end, with a route adjustment response time of ≤10 seconds. Simultaneously, the module also sets transportation time limits, with the total inter-provincial transportation time not exceeding 48 hours, and automatically issuing warnings if this time limit is exceeded.

[0076] The multi-department collaborative early warning system has three levels of warning thresholds: Level 1 warning (gas concentration exceeds national limit by 1.5 times, leak detection sensor triggered, impact acceleration ≥10g), Level 2 warning (temperature and humidity exceed standard by ±10%, route deviation ≥1 km, transportation timeout exceeds 2 hours), and Level 3 warning (gas concentration exceeds national limit by 1.0-1.5 times, temperature and humidity exceed standard by ±5%-±10%). When a transport vehicle is passing through a mountainous area in Province C, the leak detection sensor detects a waste mineral oil leak, triggering a Level 1 warning. The system automatically pushes the accident location (latitude and longitude: 35°22′36″N, 118°15′42″E), hazardous waste type (waste mineral oil), leak volume, and disposal suggestions (such as using oil-absorbing pads, setting up a warning zone) to the environmental protection departments of Provinces A, B, and C, the emergency management department, and the public security department. It also pushes the information to the transportation and disposal ends, with an emergency response time of ≤3 minutes.

[0077] (iv) Application layer deployment The regulatory platform includes national, provincial, and municipal environmental protection department monitoring terminals, utilizing both web and mobile apps. The national-level monitoring terminal allows users to view the overall situation of inter-provincial hazardous waste transportation nationwide, and to compile statistics on transportation and disposal volumes for each province. The provincial-level monitoring terminal allows users to view the real-time status of waste-generating enterprises, transport vehicles, and disposal facilities within their province, and to generate inter-provincial transportation monitoring reports. The municipal-level monitoring terminal is responsible for monitoring waste-generating enterprises and transportation origins / destination points within its jurisdiction. The monitoring platform supports cross-regional data sharing; environmental protection departments in Province A and Province B can view the location and status of transport vehicles in real time, enabling collaborative monitoring.

[0078] The enterprise side (waste-generating enterprise side and disposal agency side) is installed in the management departments of 5 waste-generating enterprises in Province A and 3 disposal agencies in Province B. The waste-generating enterprise side supports functions such as application for inter-provincial transfer of hazardous waste, online filing, and transportation tracking. The system automatically generates electronic inter-provincial transfer manifests without manual filling. The disposal agency side supports functions such as receipt confirmation, disposal process recording, and disposal result reporting. After disposal is completed, an inter-provincial disposal report is automatically generated and uploaded to the platform layer.

[0079] The transportation app is installed on the drivers' mobile phones and in-vehicle terminals of 15 transport vehicles. Drivers receive transportation tasks through the app and view the optimal route, transportation time limits, and information on checkpoints along the way. During transportation, the app displays the vehicle's location, speed, environmental parameters inside the vehicle, and leakage status in real time. In case of abnormalities, it issues an audible and visual alarm, and drivers can report to the platform and regulatory departments along the route with one click. When passing through checkpoints, drivers present electronic manifests and traceability codes through the app. After staff scan the codes to verify, they can pass through, improving traffic efficiency.

[0080] The emergency terminal is installed on the terminal equipment of national and provincial emergency rescue teams. After the first-level warning is triggered, the emergency terminal will automatically display the accident location, hazardous waste type, risk level, disposal suggestions and surrounding emergency resources (such as emergency rescue teams, oil-absorbing pad storage points, sewage treatment facilities). Commanders can use the emergency terminal to dispatch resources across regions and monitor the disposal progress in real time.

[0081] (V) Operational Results After 10 months of operation, the system in this embodiment improved the compliance rate of inter-provincial hazardous waste transportation declarations from 82% to 99.6%, the completeness rate of transportation tracks reached 100%, and the accuracy rate of abnormal warnings reached 99.0%. It successfully handled one waste mineral oil spill accident without causing environmental pollution. The average passage efficiency of transport vehicles increased by 40%, and the verification time at checkpoints was shortened from 15 minutes to 2 minutes. The efficiency of cross-regional regulatory collaboration was greatly improved, and the environmental protection departments of Province A and Province B could share data in real time, realizing closed-loop supervision of the entire process of inter-provincial hazardous waste transportation.

[0082] Comparative examples are given for Examples 1-3: Compare with Example 1-1: A conventional system lacking multi-dimensional sensor arrays and LSTM dynamic threshold calibration Configuration instructions Sensing layer: Only a single gas sensor is deployed (no temperature, humidity, liquid level / weight, or sealing status sensors) and a common GPS module (positioning accuracy ≤10 meters). There are no three-axis accelerometers or electronic seals, and no dynamic threshold calibration unit. Sensor data is uploaded directly without filtering out environmental interference.

[0083] Network layer: It adopts a single 4G communication, has no edge computing nodes, transmits data directly to the centralized server, has no blockchain consensus nodes, relies on traditional databases for data storage, and has no tamper-proof mechanism.

[0084] Platform layer: No digital twin engine, only basic data statistics function; route planning adopts fixed route navigation (without considering restricted areas or road conditions); the early warning system is a single threshold alarm with no hierarchical mechanism, and is only pushed to the enterprise manager.

[0085] Application layer: The regulatory side only supports location query and data export, while enterprises need to manually enter ledgers and there is no compliance self-inspection function.

[0086] Running effect The data accuracy rate was only 72%. Due to the lack of temperature, humidity, and sealing status detection, three hazardous waste leaks were not detected in time, and the false alarm rate of the gas sensor reached 25% (due to interference from chemical gases in the industrial park).

[0087] The data storage was found to have been tampered with by the company on two separate occasions. When trying to trace the data, no valid evidence could be provided, and the determination of responsibility took more than 72 hours.

[0088] The transportation route repeatedly passed through water source protection areas, resulting in a 30% violation rate due to the lack of intelligent planning capabilities; one leakage incident, lacking tiered early warning, resulted in an emergency response time of up to 45 minutes, with pollution spreading over an area of ​​500m². 2 .

[0089] The workload of manual verification by regulatory authorities has only decreased by 20%, and two cases of illegal transfer of hazardous waste in the park have occurred.

[0090] Core differences and weaknesses The failure to adopt the "multi-dimensional sensor group + LSTM dynamic threshold calibration" and "blockchain evidence storage" technologies of this invention results in insufficient data accuracy and traceability reliability; the lack of "intelligent path planning + hierarchical early warning" fails to meet the regulatory needs of industrial parks with multiple pollution sources and high risks, significantly increasing environmental risks.

[0091] Compare with Example 1-2: Centralized systems lacking edge computing and hybrid communication Configuration instructions Perception layer: Deploys the same multi-dimensional sensing group as in Example 1, but without edge computing nodes. Raw data is directly uploaded to the platform layer without local preprocessing function.

[0092] Network layer: Only 5G communication is used, with no redundant design for NB-IoT and satellite communication; there are no blockchain nodes, and data is stored on a single server in the park (centralized architecture).

[0093] Platform layer: It has a digital twin model, but due to the lack of edge computing preprocessing, the model update delay exceeds 300ms; the route planning does not incorporate road carrying capacity data; the early warning system only connects to the environmental protection department, without multi-departmental collaboration.

[0094] Application layer: The function is the same as in Example 1, but data transmission relies on a single 5G network and there is no backup communication solution.

[0095] Running effect Due to the lack of edge computing preprocessing, the daily data processing pressure on the platform layer increased by 3 times, the digital twin model frequently lags, and the existing prediction error reaches 15% (only 3% in Example 1).

[0096] Due to insufficient 5G signal coverage, data interruptions in the park's underground warehouse and remote transportation sections accumulated to 8 hours per month, resulting in two instances of undetected storage exceeding the expiration date.

[0097] A centralized server suffered a data loss due to a hacker attack, and 30% of the enterprises' ledger information could not be recovered; a transport vehicle collision accident resulted in a 20-minute delay in response from the public security and emergency departments due to the lack of multi-departmental coordinated early warning.

[0098] The optimization rate of hazardous waste disposal process is only 10% (40% in Example 1), and the overall regulatory efficiency of the park has improved by less than 30%.

[0099] Core differences and weaknesses The failure to adopt the "edge computing + hybrid communication redundancy design" and "multi-department collaborative early warning" technologies of this invention resulted in insufficient data transmission stability and processing efficiency, poor emergency response coordination, and an inability to adapt to the complex network environment and emergency needs of industrial parks.

[0100] Compare with Examples 1-3: The underlying IoT system lacking digital twins and blockchain traceability Configuration instructions Perception layer: Deploys a multi-dimensional sensor group with dynamic threshold calibration function. The sensor configuration is the same as in Example 1.

[0101] Network layer: It adopts 5G+NB-IoT hybrid communication, with no edge computing nodes and data directly uploaded; there are no blockchain consensus nodes, and data is stored on cloud servers (centralized).

[0102] Platform layer: No digital twin engine, only data visualization function; route planning is based on basic GPS navigation (without considering the characteristics of hazardous waste); the early warning system is a level 2 early warning system with no risk simulation function.

[0103] Application layer: The regulatory side only supports real-time data viewing, without functions for stock prediction or risk simulation; the enterprise side does not have an automatic ledger generation function.

[0104] Running effect Due to the lack of a digital twin model, it was impossible to predict the risk of storage warehouse capacity saturation, resulting in three instances of hazardous waste exceeding its expiration period, with the longest exceeding the period by 7 days.

[0105] The transportation route failed to consider the characteristics of hazardous waste (such as the risk of corrosion from waste acid), and twice passed through sections of bridges with insufficient load-bearing capacity, causing safety hazards; the route deviation warning response time reached 30 seconds (only 8 seconds in Example 1).

[0106] Centralized storage resulted in one instance of a company tampering with transportation trajectory data, making it impossible to verify responsibility during tracing, and the dispute resolution process took 5 days; the data query response time reached 5 seconds (compared to only 2 seconds in Example 1).

[0107] Regulatory authorities were unable to anticipate risks in advance, and the number of incidents requiring reactive handling increased by 4 compared to Example 1, resulting in an environmental risk prevention and control efficiency of less than 50%.

[0108] Core differences and weaknesses The failure to adopt the "digital twin engine" and "blockchain consortium chain traceability" technologies of this invention results in a lack of predictive regulatory capabilities, insufficient traceability reliability, and an inability to achieve closed-loop management of hazardous waste throughout its entire life cycle in industrial parks.

[0109] Compare with Example 2-1: Traditional surveillance systems lacking infectious risk sensors and encryption modules Configuration instructions Sensing layer: Only weight sensors and ordinary temperature and humidity sensors are deployed, without infection risk sensors, sealing status sensors and sharp object counting sensors; GPS positioning accuracy ≤ 8 meters, without dynamic threshold calibration function.

[0110] Network layer: It adopts LAN + 4G communication, has no blockchain nodes, transmits data in plaintext, and stores it on the hospital's local server (without backup); it has no edge computing nodes, and data is uploaded directly.

[0111] Platform layer: No digital twin engine, only simple data statistics function; no intelligent route planning, transportation routes are selected by the driver; the early warning system only targets excessive temperature and humidity, and does not provide early warning of infectious risk.

[0112] Application layer: The regulatory end only supports data export and has no dedicated monitoring module; the hospital end requires manual filling of ledgers and has no HIS system integration function; the treatment end has no electronic confirmation process.

[0113] Running effect Due to the lack of infectious risk sensors, two instances of excessive bacterial concentrations in medical waste from the infectious disease department went undetected, leading to an increased risk of cross-contamination in the temporary storage area and an infectious waste disposal compliance rate of only 85% (compared to 99.5% in Example 2).

[0114] Plaintext data transmission resulted in one instance of indirect leakage of patient privacy information; a malfunction of the hospital's local server caused the loss of three days' worth of ledger data, making it impossible to trace the flow of 20 batches of medical waste.

[0115] The transportation route repeatedly passed through the vicinity of schools (restricted areas), with a violation rate of 25%; one instance of seal damage went undetected; and the emergency response time after a medical waste leak was 50 minutes, involving two residential areas.

[0116] The error rate of manual ledger entry reached 10%, the workload of manual verification by regulatory authorities was reduced by only 15%, and the average temporary storage time of medical waste was as long as 36 hours (only 24 hours in Example 2).

[0117] Core differences and weaknesses The failure to adopt the "infectious risk sensor + SM4 / SM2 encryption" and "multi-dimensional status monitoring" technologies of this invention results in the inability to control core risks (infection, privacy leakage) in medical scenarios; the lack of intelligent planning and early warning does not meet the requirements of high safety and high traceability for medical hazardous waste.

[0118] Compare with Example 2-2: A semi-automated system lacking blockchain traceability and electronic seals Configuration instructions Perception layer: Deploys multi-dimensional sensor groups (including infectious risk sensors) with dynamic threshold calibration function; however, it uses ordinary physical seals (without electronic records) and the sharps counting sensor has low accuracy (error ±5).

[0119] Network layer: It adopts 5G+LAN communication, has no blockchain nodes, and stores data in the provincial government cloud (centralized); it has edge computing nodes, but no data preprocessing optimization (only simple filtering).

[0120] Platform layer: It has a digital twin model, but it does not access infectious risk data and has no transmission path simulation function; the path planning does not have a time window limit; the early warning system only pushes to hospitals and does not have multi-department collaboration.

[0121] Application layer: The regulatory end supports real-time monitoring, but lacks traceability coding query function; the hospital end can automatically generate ledgers, but lacks compliance self-inspection; the transportation end lacks electronic manifest function.

[0122] Running effect Due to the lack of blockchain traceability, the transportation trajectory records for three batches of infectious waste were incomplete, making it impossible to verify whether they passed through illegal areas, and the determination of responsibility took 48 hours; the ordinary seals were tampered with twice, posing a risk of illegal transfer.

[0123] Errors in sharps counting led to an 8% error rate in the classification and statistics of medical waste, resulting in multiple discrepancies in the quantities received by disposal facilities; the digital twin model, lacking infectious data, had a risk simulation accuracy of only 60% (compared to 95% in Example 2).

[0124] The transport vehicles repeatedly made trips during peak daytime hours, resulting in a complaint rate of 15%; one instance of a bacterial concentration exceeding the standard was only sent to the hospital, and the emergency response department failed to respond in a timely manner, increasing the risk of infection spread.

[0125] The average time to generate a medical waste disposal report is 24 hours (only 6 hours in Example 2), and the efficiency of cross-hospital regulatory data sharing is less than 40%.

[0126] Core differences and weaknesses The failure to adopt the "blockchain consortium chain traceability + electronic seal" and "multi-department collaborative early warning" technologies of this invention has resulted in a broken traceability chain for medical hazardous waste, untimely emergency response, and an inability to meet the collaborative supervision needs across hospitals and departments.

[0127] Compare with Example 2-3: The lack of a basic IoT system for digital twins and tiered early warning. Configuration instructions Perception layer: Deploy a complete multi-dimensional sensor group with dynamic threshold calibration and electronic sealing functions. The sensor configuration is the same as in Example 2.

[0128] Network layer: It adopts 5G + local area network + blockchain consortium chain, and the data transmission and storage security meet the requirements; there are no edge computing nodes, and the data is directly uploaded to the platform.

[0129] Platform layer: No digital twin engine, only data visualization function; route planning is integrated into restricted areas, but there is no time window or emergency rescue point planning; the early warning system is a single level, without differentiated handling suggestions.

[0130] Application layer: The regulatory side supports traceability and query, but there is no dedicated monitoring; the hospital side has compliance self-inspection, but there is no HIS system integration; the emergency response side has no resource scheduling function.

[0131] Running effect Due to the lack of a digital twin model, it was impossible to predict the capacity of temporary storage rooms in each hospital. Medical waste accumulated beyond its expiration period four times, with the longest overdue period reaching 12 hours. The transmission path of infectious waste could not be simulated, and the risk assessment took up to 30 minutes (compared to only 5 minutes in Example 2).

[0132] The transportation route lacked planned emergency rescue points, resulting in a 15-minute delay in the allocation of rescue supplies and a 40% reduction in response efficiency after a single incident of seal breach. Furthermore, the absence of time window restrictions meant that 30% of transportation tasks occurred during peak hours, doubling the duration of congestion.

[0133] The single-level early warning system leads to a large amount of regulatory resources being consumed in non-emergency situations (such as slight exceedances of temperature and humidity), with an effective early warning response rate of only 70%. The emergency response system lacks resource scheduling capabilities, and in one infectious leak incident, the surrounding disinfection supplies were not allocated in a timely manner.

[0134] The hospital's records and HIS system data cannot be synchronized, the workload of manual verification is still large, and the efficiency of supervision is improved by less than 50% (80% in Example 2).

[0135] Core differences and weaknesses The failure to adopt the "digital twin engine + infectious risk simulation" and "three-level graded early warning + differentiated disposal" technologies of this invention has resulted in a lack of predictive regulatory capabilities, insufficient accuracy of early warning and emergency response efficiency, and an inability to meet the high-risk and high-timeliness regulatory needs of medical hazardous waste.

[0136] Compare with Example 3-1: A single network system lacking hybrid communication and satellite communication Configuration instructions Perception layer: Deploys multi-dimensional sensor groups (including leak detection sensors) with dynamic threshold calibration function; however, GPS positioning accuracy is ≤5 meters, there is no Beidou dual-mode redundancy, and the accuracy of the three-axis accelerometer is low (±0.1g).

[0137] Network layer: Only 4G communication is used, without 5G, NB-IoT and satellite communication modules; it has edge computing nodes, but no local caching function; the blockchain nodes only include waste-generating provinces and disposal provinces, and there are no backup nodes in transit provinces.

[0138] Platform layer: It has a digital twin model, but due to communication interruption, dynamic data updates are discontinuous; the route planning is based on the basic GIS map and lacks meteorological and road load data; the early warning system does not have cross-province push function.

[0139] Application layer: The regulatory end only supports data querying of the waste-generating province and the disposal province, and has no access to the transit provinces; the transportation end does not have electronic manifest function, and paper documents are required; the emergency end does not have cross-regional resource dispatch function.

[0140] Running effect When transporting hazardous waste across provinces through mountainous and remote areas, the cumulative duration of 4G signal interruption reached 12 hours per month, resulting in the loss of transport tracks for three batches of hazardous waste, making it impossible to verify whether they deviated from the route; the lack of satellite communication led to a complete data interruption during a snowstorm, resulting in a regulatory vacuum of up to 4 hours.

[0141] The blockchain lacks provincial nodes, and when it passes through a provincial checkpoint for verification once, real-time data cannot be retrieved, extending the passage time to 30 minutes (compared to only 2 minutes in Example 3); the positioning accuracy is insufficient, and there were two misjudgments of path deviation, affecting transportation efficiency.

[0142] The route planning failed to consider blizzard conditions and bridge load-bearing capacity. On one occasion, a transport vehicle skidded on an icy section of road, and the lack of real-time route adjustment resulted in an 8-hour delay. Additionally, due to communication interruptions, a leak detection sensor failed to report a leak in a timely manner, allowing the contamination area to expand to 1000m. 2 .

[0143] Cross-provincial early warning pushes are delayed by up to 1 hour, emergency resource dispatch requires manual coordination, and the response time is as long as 60 minutes (only 3 minutes in Example 3). The efficiency of cross-provincial regulatory collaboration is less than 30%.

[0144] Core differences and weaknesses The failure to adopt the "5G+NB-IoT+satellite communication hybrid networking" and "cross-regional blockchain alliance chain" technologies of this invention has resulted in insufficient continuity of cross-provincial transportation data transmission and cross-departmental collaboration, failing to address the pain points of regulatory vacuum in remote road sections and low efficiency of cross-regional verification.

[0145] Compare with Example 3-2: Traditional transportation systems lacking intelligent route planning and cross-regional collaboration Configuration instructions Perception layer: Deploys multi-dimensional sensor groups with dynamic threshold calibration and leak detection functions; satellite communication module is fully configured with positioning accuracy ≤3 meters.

[0146] Network layer: It adopts a hybrid communication architecture with edge computing nodes and blockchain consortium chains (including nodes in transit provinces); however, data encryption only uses the ordinary AES algorithm and does not have the SM4 / SM2 national cryptographic standard.

[0147] Platform layer: It has a digital twin model, but the path planning adopts a fixed route (without dynamic adjustment); the early warning system only pushes to the regulatory departments of the waste-generating province and the disposal province, without linkage with the emergency departments of the provinces through which it passes; there is no function to limit transportation time.

[0148] Application layer: The regulatory end supports cross-provincial data query, but lacks a collaborative regulatory interface; the transportation end has electronic manifests, but lacks real-time road condition push notifications; the emergency end lacks personalized handling suggestions.

[0149] Running effect The route planning lacked dynamic adjustment capabilities, resulting in five instances of transportation delays due to road construction and congestion, with the longest delay reaching 12 hours. It also failed to avoid nature reserves, leading to three violations and penalties from the provinces it passed through.

[0150] The lack of national cryptographic encryption led to the interception of a cross-provincial data transmission, resulting in the leakage of hazardous waste composition information; the transportation time was unlimited, with one batch of waste mineral oil being transported for as long as 72 hours, exceeding the safe transportation period and increasing the risk of deterioration.

[0151] The early warning system lacked coordination with emergency response departments in the provinces along the route. After one leak occurred, the emergency response departments in the provinces along the route failed to respond in a timely manner, and the arrival of rescue supplies was delayed by 25 minutes. There were no personalized handling suggestions, and the personnel handling the incident used a general solution, which reduced the efficiency of leak cleanup by 50%.

[0152] The lack of a cross-provincial collaborative supervision interface means that the waste-generating province and the disposal province cannot synchronize the transportation status in real time, which increases the cost of information communication and improves supervision efficiency by less than 40% (75% in Example 3).

[0153] Core differences and weaknesses The failure to adopt the "improved A* algorithm intelligent path planning" and "multi-province collaborative early warning + national cryptographic encryption" technologies of this invention results in low efficiency of cross-provincial transportation, high data security risks, and insufficient targeted emergency response, failing to meet the regulatory requirements for long-distance cross-regional transportation.

[0154] Compare with Example 3-3: Centralized Transportation Systems Lacking Edge Computing and Digital Twin Prediction Configuration instructions Perception layer: Deploys multi-dimensional sensor groups with hybrid communication and blockchain consortium chain functions; but has no edge computing nodes, and raw data is directly uploaded to the platform layer; has no leak detection sensors (relies only on gas sensors).

[0155] Network layer: The communication and blockchain configuration is the same as in Example 3, but there is no local caching function, and data transmission relies on cloud storage.

[0156] Platform layer: No digital twin engine, only trajectory playback function; route planning incorporates multiple factors, but no transportation time prediction function; the early warning system is a level 2 early warning system, without risk level assessment.

[0157] Application layer: The regulatory end lacks national overall data statistics function; the enterprise end lacks the function of automatically generating electronic manifests for cross-provincial transfers; the emergency end lacks the function of visualizing the distribution of surrounding resources.

[0158] Running effect Without edge computing preprocessing, the platform layer data processing latency reaches 200ms, and the path adjustment response time reaches 30 seconds (only 10 seconds in Example 3); without leak detection sensors, a liquid leak caused a 15-minute delay in alarm due to the gas sensor not triggering in time.

[0159] Without digital twin prediction, transportation time could not be predicted, and there were three instances of delays in receiving the waste due to timeouts, resulting in insufficient storage capacity. Track playback only supports post-event queries and cannot intervene in violations in real time, with two instances of illegal stays.

[0160] The Level 2 warning lacked a risk assessment, and a minor temperature and humidity exceedance was mistakenly identified as an emergency warning, consuming a large amount of emergency resources; the emergency response lacked resource visualization, making it impossible for rescuers to quickly locate the oil-absorbing pad storage points, extending the response time by 30 minutes.

[0161] Enterprises need to manually fill out inter-provincial transfer forms, with an error rate of 12%, and the approval process takes an average of 48 hours (only 6 hours in Example 3); there is a lack of overall national data statistics, and national regulatory authorities cannot grasp the overall situation of inter-provincial transportation.

[0162] Core differences and weaknesses The failure to adopt the "edge computing + local caching" and "digital twin prediction + risk level assessment" technologies of this invention results in low data processing efficiency, lack of predictive supervision, and inability to achieve multi-level collaborative management and control at the national, provincial, and transit provinces, which does not meet the overall regulatory requirements for cross-provincial transportation.

[0163] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of protection claimed by the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A hazardous waste monitoring system based on the Internet of Things, characterized in that, It includes the perception layer, network layer, platform layer, and application layer; The sensing layer includes a multi-dimensional sensor group deployed at hazardous waste generation sites, storage warehouses, transport vehicles, and disposal terminals. The multi-dimensional sensor group includes a gas sensor, a temperature and humidity sensor, a liquid level / weight sensor, a GPS / BeiDou positioning module, a three-axis accelerometer, and a sealing status sensor. It is used to collect physicochemical parameters, location information, and status data of hazardous waste throughout its entire life cycle. The sensing layer also has a built-in dynamic threshold calibration unit that performs real-time calibration of the sensor data based on the LSTM algorithm. The network layer includes edge computing nodes, a hybrid communication module, and blockchain consensus nodes. The edge computing nodes preprocess the sensor data and perform initial anomaly detection. The hybrid communication module adopts a redundant design of 5G, NB-IoT, and satellite communication. The blockchain consensus nodes include regulatory department nodes, enterprise nodes, and third-party testing nodes to realize data on-chain evidence storage. The platform layer includes a digital twin engine, a blockchain traceability module, an intelligent route planning module, and a multi-department collaborative early warning system. The digital twin engine constructs a digital twin model of the entire life cycle of hazardous waste. The blockchain traceability module achieves tamper-proof data traceability based on a consortium blockchain architecture. The intelligent route planning module combines GIS maps and real-time traffic conditions to generate the optimal transportation route. The multi-department collaborative early warning system automatically matches the responsible parties and triggers tiered early warnings. The application layer includes a regulatory terminal, an enterprise terminal, a transportation terminal, and an emergency terminal. Each terminal interacts with data and calls functions through the platform layer, forming a closed-loop supervision process.

2. The hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The calibration process of the dynamic threshold calibration unit is as follows: collect sensor data and environmental interference parameters, establish a sensor data-interference parameter mapping model through the LSTM algorithm, adjust the sensor threshold range in real time, and eliminate abnormal interference data.

3. The hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The blockchain consensus nodes adopt the PBFT consensus mechanism. The block data includes hazardous waste classification information, sensor data, transportation trajectory data, disposal result data, and signature information of responsible parties at each stage. The block generation interval does not exceed 30 seconds.

4. The hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The digital twin engine acquires physical scene data through 3D laser scanning, constructs a static model by combining it with BIM technology, and updates the dynamic model in real time by accessing data from the perception layer, thereby enabling the prediction of hazardous waste inventory, simulation of risk points, and optimization of disposal processes.

5. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The intelligent route planning module is based on the A* algorithm and incorporates data on hazardous waste transportation restricted areas, road carrying capacity, and real-time traffic congestion to dynamically adjust transportation routes. The route deviation warning response time does not exceed 10 seconds.

6. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The multi-department collaborative early warning system is set up with a three-level early warning mechanism. The first-level early warning triggers a joint action among environmental protection, emergency response, and public security departments. The second-level early warning is pushed to the regulatory departments in the jurisdiction and the enterprise leaders. The third-level early warning only sends rectification reminders to enterprises. The early warning information includes the location and type of risk and disposal suggestions.

7. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The multi-dimensional sensing group on the transport vehicle also includes a video monitoring module and an electronic seal sensor. The electronic seal sensor is linked to the door lock of the transport vehicle, and automatically triggers positioning tracking and video recording when the seal is abnormal.

8. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The platform layer also includes a data encryption module, which uses the SM4 symmetric encryption algorithm to encrypt transmitted data and the SM2 asymmetric encryption algorithm to authenticate node identities.

9. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The enterprise side of the application layer supports online declaration of hazardous waste, automatic generation of ledgers, and self-inspection of compliance, and automatically links to the updated classification standards of the national hazardous waste list.

10. A hazardous waste monitoring system based on the Internet of Things according to claim 1, characterized in that: The sensors in the sensing layer are designed to be waterproof, corrosion-resistant, and explosion-proof, with a protection level of no less than IP67, making them suitable for the harsh environments of hazardous waste storage and transportation.

Images